The relationship between the Atlantic meridional overturning circulation and coastal sea level in CMIP5 models

Student thesis: Master's ThesisMaster of Science by Research (MScR)

Abstract

The importance of the Atlantic meridional overturning circulation (AMOC) for the Earth’s climate is widely recognised. As a complement to temporally and spatially scarce in-situ AMOC observations, several studies have sought a connection between AMOC and sea level variability. Tide gauges and coastal sea level (CSL) are of special interest as they offer over century-long data and cost-effective monitoring. Yet, to what extent AMOC variability influences CSL is subject to conflicting evidence, challenging the validity of CSL as AMOC proxies.
This study examines the relationship between the AMOC and sea level variability along the North American east coast in a large number (>40) of CMIP5 models over the period 1920-2300 on interannual and decadal timescales. As such it extends existing studies having focused primarily on a single ocean model, observations or a limited subset of state-of-the-art coupled models. The spatiotemporal characteristics of the relationship are examined by; 1) assessing the direct relationship, including a comparison between initial AMOC simulations and a sea surface height-based reconstruction of the AMOC derived from linear regression, and 2) calculating an AMOC index based on alongshore CSL composites. About half of the CMIP5 models generate moderate to strong AMOC-CSL anticorrelations (<-0.5) along large latitudinal stretches, particularly between 35°N-48°N, on decadal timescales in both historical simulations and 21st century projections. Concurrently, if extending the analysis up to year 2300, a noticeable intra-model spread is noted in the magnitude of the relationship over the four different 100-yr intervals (1900-2300), suggesting that the AMOC-CSL relationship may be non-stationary and dependent on the temporal period examined. The models generally underestimate both the observed AMOC and CSL variability. The implications of these results are discussed with a focus on the suitability of CSL-based AMOC proxies, reliability of CMIP5 projections and future CSL changes.   
Date of Award12 May 2020
Original languageEnglish
Awarding Institution
  • University of Bristol
SupervisorRory J Bingham (Supervisor) & Paul J Valdes (Supervisor)

Cite this

'